OBJECTIVETo evaluate the performance of 18F-FDG three-head tomography with coincidence imaging and serum tumor marker assays in identifying lung lesions in 104 patients with abnormal findings on chest X-ray or computer tomography.

METHODSA prospective evaluation of 18F-FDG coincidence imaging and the measurement of 3 serum markers for lung cancer ( carcinoembryonic antigen, CYFRA21-1 and neuron specific enolase) were performed within one week in 104 inpatients with suspected lung malignancy. All images were analyzed visually. It was considered positive for malignancy if the 18F-FDG uptake was increased relative to that in the adjacent lung tissue, and was focal. The serum tumor marker test was considered positive for malignancy if the serum level of at least one marker was elevated.

RESULTS66 patients were proven to have lung cancer by pathology, and 38 patients had benign lung diseases. The sensitivity, specificity, accuracy of 18F-FDG coincidence imaging and serum tumor markers in assessing lung cancers were 80. 0% , 77. 2% , 77. 9% and 56. 0% , 60. 9%, 64. 4% , respectively. 18F-FDG coincidence images in assessing lung lesions showed significantly higher sensitivity, specificity and accuracy than serum tumor markers. Four patients with lung cancer had negative findings on 18F-FDG coincidence images but showed positive serum markers.

CONCLUSION18F-FDG coincidence imaging is a powerful tool for evaluating patients with lung lesions suggestive of malignancy. Although the determination of serum marker levels is less accurate than 18F-FDG coincidence imaging, the combination of a positive 18F-FDG coincidence result and positive tumor markers may be helpful in improving the diagnosis of lung cancers.